Monitoring Land Degradation Using Drone Photogrammetry for the Mugoboka (Eastern Bujumbura, Burundi) Case Study

Previously published studies on soil loss by erosion in Burundi have focused on sheet and rill erosion. However, other forms of erosion such as gullying and landslides should not be neglected in mountainous countries like Burundi, and also have an effect on soil health. Recent developments have demonstrated the potential of drone remote sensing in estimating several factors and forms of soil erosion, particularly in terms of cost effectiveness. However, such applications are non-existent in Burundi. In this work, we evaluate the relevance of drone photogrammetry for soil erosion assessment and modelling in the Mugoboka area (Eastern Bujumbura, Burundi). This region is facing severe land degradation, as a result of different forms of erosion and demographic pressure with mismanaged house constructions and crop production. Given its proximity to the University of Burundi, this area is the most frequently monitored by drone photogrammetry for training and research purposes since June 2022. With a low-cost drone system (DJI Mavic 2 Pro and DJI Mavic 3 Enterprise with D-RTK 2 module) we started investigating the spatio-temporal evolution of gully erosion, landslides and built-up areas at Mugoboka from June 2022 to February 2024. Drone photogrammetry was completed using Agisoft Metashape software and the resulting orthomosaics and Digital Terrain Models (DTMs) were further processed in QGIS. So far, five overlapping zones have been investigated during this time span and the results obtained show an alarming situation of land degradation and land mismanagement in Mugoboka area, including ongoing gullying and house construction on landslide-prone sites. The quality of drone-derived Orthomosaics and DTMs is encouraging in terms of detection of small topographic and land use changes. The detailed DTMs can be used to precisely locate surface water flow paths and concentrations that can in turn be accounted for in designing protective measures. We conclude that drone remote sensing offers promising results for the holistic and low-cost monitoring and modelling of soil loss by erosion in Burundi.